Murray Straus

Available online at www.sciencedirect.
com
Children and Youth Services Review 30 (2008) 252 – 275

www.elsevier.com/locate/childyouth
Dominance and symmetry in partner violence by male and

female university students in 32 nations
Murray A. Straus ⁎
Family Research Laboratory, University of New Hampshire, Durham, NH 03824, United States
Available online 13 October 2007
Abstract
The study investigated the widely held beliefs that physical violence against partners (PV) in marital, cohabiting, and dating
relationships is almost entirely perpetrated by men, and that the major risk factor for PV is male dominance in the relationship. The
empirical data on these issues were provided by 13,601 university students in 32 nations who participated in the International
Dating Violence Study. The results in the first part of this paper show that almost one-third of the female as well as male students
physically assaulted a dating partner in the previous 12 months, and that the most frequent pattern was bidirectional, i.e., both were
violent, followed by “female-only” violence. Violence by only the male partner was the least frequent pattern according to both
male and female participants. The second part of the article focuses on whether there is gender symmetry in a crucial aspect of the
etiology of partner PV — dominance by one partner. The results show that dominance by either the male or the female partner is
associated with an increased probability of violence. These results, in combination with results from many other studies, call into
question the assumption that PV is primarily a male crime and that, when women are violent, it is usually in self-defense. Because
these assumptions are crucial elements in almost all partner PV prevention and treatment programs, a fundamental revision is
needed to bring these programs into alignment with the empirical data. Prevention and treatment of PV could become more
effective if the programs recognize that most PV is bidirectional and act on the high rate of perpetration by women and the fact that
dominance by the female partner is as strongly related to PV as dominance by the male partner.
© 2007 Elsevier Ltd. All rights reserved.
Keywords: Domestic violence; Gender; Prevention; Assault; Patriarchy
1. Introduction
This article reports the results of an empirical investigation of two of the most controversial and important issues in
understanding physical violence between partners in marital, cohabiting, or dating relationships. The answers to these
questions can have profound implications for prevention and treatment of partner violence.
1. Is partner violence primarily perpetrated by men, as compared to women, and as compared to both partners
engaging in physical violence?
⁎ Tel.: +1 603 862 2594.

E-mail address: murray.straus@unh.edu.
URL: http://pubpages.unh.edu/~mas2.
0190-7409/$ - see front matter © 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.childyouth.2007.10.004
M.A. Straus / Children and Youth Services Review 30 (2008) 252–275 253
2. To what extent is dominance by the male partner associated with partner violence, as compared to dominance by the
female partner? In short is the risk factor male dominance or dominance by one partner, regardless of whether it is
the male female partner?
For reasons explained elsewhere (Straus, 2007b), mentioning these two issues as topics for empirical investigation is
often regarded as undermining efforts to end partner violence (PV from here on), and often greeted with hostility
because they implicitly challenge two core principles that underlie most efforts to prevent and treat PV.
The first principle is that PV is primarily perpetrated by men: In an article on “Sexual Inequality, Cultural Norms,
and Wife-Beating” published 30 years ago (Straus, 1976), I stated that “wives are much more often the victim of
violence by their husbands than the reverse”. In relation to the second principle, in that article I attributed male PV to
“the hierarchical and male-dominant nature of society…” The second core principle is that when men are violent the
purpose is to coerce and dominate, whereas when women are violent it is almost always an act of self-defense or a
response to unbearably humiliating and dominating behavior by the male partner. The idea that women are motivated to
hit in order to coerce a male partner, or out of rage and anger over misbehavior by a male partner (such as sexual
infidelity), is regarded as outrageous, and is taken as a sign of sexism and misogyny.
In the 35 years since I began research on PV, bit by bit, these assumptions about prevalence and etiology have been
contradicted by a mass of empirical evidence from my own research and from research by many others. Consequently, there is a
need for a much more multi-faceted view of PV. This would recognize the overwhelming evidence that women assault their
partners at about the same rate as men, and that the motives for violence by both males and females are diverse. However, until
recently, few have accepted this evidence, and some of those few will not publicly express their position for fear of the type of
ostracism to which it will expose them (Straus, 1990c, in press). Instead, the evidence on gender symmetry in prevalence and
etiology is typically disregarded and often explicitly denied, or withheld from publication (Straus, 2007b, in press). As will be
suggested in the conclusion, this denial has crippled prevention and treatment efforts.
The focus of this paper is on physical assault because that is the aspect of partner maltreatment that has been the focus
of the most controversy. In the context of this paper, PV refers to physical assault. Two aspects of gender symmetry in PV
will be addressed: bidirectional perpetration by men and women and parallel etiology of violence. The main objectives
of this article are to present the results of a cross-national study of these two aspects of gender symmetry and to draw out
their implications for prevention and treatment programs. An additional objective is to illustrate the use of an easily
applied typology. This classifies cases into Male-Only violence, Female-Only violence, and Both Violent. Use of these
simple but crucial categories are needed to help research and prevention and treatment programs act on the implications
for prevention and treatment which flow from the empirical results presented in this article.
There are many reasons for the conceptualization of PV as a problem of violence against women, some of which will
be mentioned in this article, and are presented more fully in Straus (2007b). One of the most important is that the injury
rate for male perpetrated violence is much higher than the rate for physical attacks by women. For this reason and
because of the lesser financial resources of women, there is a much greater need for victim services for female than for
male victims of PV.
2. Previous evidence on bidirectional perpetration
The importance of data on bidirectionality is based on the assumption that violence occurs in the context of an
ongoing system of family relationships (Winstok, 2007). To the extent that this is the case, research and clinical work
on PV needs to take into account the behavior of both partners in the family system, including violence by both
partners. This applies even when it might seem that only information on the behavior of one of the partners is needed,
such as measuring progress in a treatment program for male batterers. Research has shown that the cessation of
violence by one partner is highly dependent on whether the other partner also stops hitting (Feld & Straus, 1989; Gelles
& Straus, 1988). Thus, even when monitoring a treatment program for a designated perpetrator, it is crucial to know the
extent to which the partner has also ceased acts of physical aggression.
Several studies, including large and nationally representative samples, have found that Female-Only violence is as
prevalent as or more prevalent than Male-Only violence, and that the most prevalent pattern is bidirectional violence;
i.e., both partners are violent. The 1975 and the 1985 National Family Violence Surveys found that about half of the
violence was bidirectional, one-quarter was male-only, and one-quarter was female-only (Gelles & Straus, 1988;
Straus, Gelles, & Steinmetz, 1980, 2006). The National Comorbidity Study (Kessler, Molnar, Feurer, & Appelbaum,
254 M.A. Straus / Children and Youth Services Review 30 (2008) 252–275
2001) found similar percentages. There are many other studies showing bidirectionality (Anderson, 2002; Capaldi &
Owen, 2001; McCarroll, Ursano, Fan, & Newby, 2004; Moffitt, Caspi, Rutter, & Silva, 2001; Whitaker, Haileyesus,
Swahn, & Saltzman, 2007; Williams & Frieze, 2005).
A limitation of the studies cited is that the data come from general population samples rather than from “clinical”
samples such as clients of shelters for battered women or men arrested for PV. At that “clinical level” of PV, male
predominance may prevail or may be more common than in general population violent couples. However, because the
informal rules of most organizations serving female victims forbid asking women about their own violence, data is scarce.
Nevertheless, bidirectionality has been found in a number of studies, starting with the Walker's Battered Woman
Syndrome. Walker found that 1 out of 4 women in battering relationships had “used physical force to get something you
wanted” (Walker, 1984, p. 174). Giles-Sims (1983 #509) found that in the year prior to coming to a shelter, 50% of the
women reported assaulting their partner, and in the six months after leaving the shelter, 42% reported assaulting a partner.
Moreover, Giles-Sims's case study data suggest that these assaults were not likely to have been in self-defense. A study by
Gondolf of men in four batterer treatment programs found that 69% of the female partners had also assaulted their partner
(Gondolf, 2002, p. 104). Thus, although there are important differences between clinical samples and general population
samples of violent couples, in respect to bidirectionality, they may not be as great as I previously thought (Straus, 1990b).
Analyses of couples in the Dunedin longitudinal cohort study found that “The less severe form (of PV) involved primarily
woman-to-man abuse, but the clinical form involved abuse practiced by both men and women” (Ehrensaft, Moffitt, &
Caspi, 2004). Thus, although there are large numbers of women who are victims of repeated severe injury-producing
violence whose only violence is in self-defense, and who urgently need assistance, the empirical evidence indicates that
mutual violence (both partners violent) is the predominant pattern at all levels of violence.
3. Research on gender symmetry in dominance
3.1. Symmetry in dominance
The scholarly literature on PV contains hundreds of publications which attribute PV as a method used by men to
maintain dominance in the relationship. One of my articles (Straus, 1976) is an early example. A recent and highly
influential example is the World Health Organization report on violence (Krug, Dahlberg, Mercy, Zwi, & Lozano,
2002). Another recent example by an experienced and respected researcher is the assertion by Hamberger (Hamberger
& Guse, 2002) that “Men in contrast {to women} appear to be motivated to use violence to dominate and control”.
Their article cited about 80 studies, but none of them provided empirical evidence on gender differences in dominance
and control motivation. This is not because there are no studies comparing relationship dominance by men and women.
There are a number of such studies (Ehrensaft & Vivian, 1999 #10997; Felson & Outlaw, 2007; Laroche, 2005 #8553;
Oswald & Russell, 2006 #11180; Stets, 1991 #10983; Stets & Hammons, 2002 #10742). All found no important
difference in control or dominance by male and female partner.
3.2. Dominance and violence
As for gender differences in the link between dominance and partner violence, the study of a nationally representative
sample of 2143 American couples reported in Behind Closed Doors: Violence In The American Family (Straus et al., 1980,
2006) found that equalitarian couples were the least violent and that dominance by either the female or male partner was
associated with an increased rate of violence. Since then, several other studies have found that dominance or control by
women as well as men is associated with an increased rate of violence by women (Kim & Emery, 2003; Medeiros & Straus,
2006; Stets & Pirog-Good, 1990; Straus et al., 1980, (2006); Straus & Members of the International Dating Violence
Research Consortium, 2006; Sugihara & Warner, 2002; Tang, 1999). These results suggest that whenever there is
dominance of one partner, there is an increased risk of violence by the dominant partner to maintain the dominant position,
or by the subordinate partner to achieve something blocked by the dominant partner, or to change the power structure.
3.3. Self-defense
Articles asserting the male dominance theory of PV usually also assert that, when women are violent, it is primarily
in self-defense, or in response to years of brutal violence and subjugation. This follows from the beliefs that it is men
who perpetrate PV and that men, but not women, assault to achieve dominance and control. If neither of those beliefs is
correct, perhaps the idea that women's violence is primarily an act of self-defense is also not correct. This is an
important aspect of PV that needs to be clarified, and is also important for the bidirectionality data to be presented
because it might explain predominance of bidirectional violence.
On this issue as well as the two core issues, there is a huge discrepancy between the assertions and the evidence. For
example, the influential World Health Organization report on violence states that “Where violence by women occurs it
is more likely to be in the form of self-defense (32, 37, 38)” (Krug et al., 2002). However, examination of references 32,
37, and 38 found that although all three asserted that women's violence was primarily in self-defense, #32 Saunders
(Saunders, 1986) reported no data on self-defense, #37 DeKesseredy et al. (DeKeseredy, Saunders, Schwartz, &
Shahid, 1997) does report data but their data shows that only 6.9% of the women acted in self-defense, and reference
#38 Johnson & Ferraro (Michael P. Johnson & Ferraro, 2000) is a review article that cites references 32 and 37 and
other references which also present no empirical data.
At least six other studies report data on self-defense. Like the DeKesseredy et al. study, five of the six found that
only a small percentage of female violence was in self-defense (Carrado, George, Loxam, Jones, & Templar, 1996;
Cascardi & Vivian, 1995; Felson & Messner, 1998; Follingstad, Wright, Lloyd, & Sebastian, 1991; Pearson, 1997;
Sarantakos, 1998; Sommer, 1996). For the one study that found high rates of self-defense, the percentage in self-
defense was slightly greater for men (56%) than for women (42%) (Harned, 2001). Rather than self-defense, the most
usual motivations for violence by women, like the motivations of men, are coercion, anger, and punishing misbehavior
by their partner (Cascardi & Vivian, 1995; Fiebert & Gonzalez, 1997). For example, Pearson (1997) reports that 90% of
the women she studied assaulted their partner because they were furious or jealous, or frustrated and not because they
tried to defend themselves. These motives are parallel to the motivations of male perpetrators. Research on homicides
by women shows similar results. For example, Jurik and Gregware (1989) studied 24 female perpetrated homicides and
found that 60% had a pervious criminal record, 60% had initiated use of physical force, and only 21% of the homicides
were in response to “prior abuse” or “threat of abuse/death”.
3.4. Conceptualization and measurement of dominance
This study used the conceptualization of dominance based on the feminist literature, and the parallel measure of
dominance, developed by Hamby (Hamby, 1996). This identifies three different forms or modalities of dominance. Each
represents a departure from an equalitarian relationship. Authority is a relationship in which one partner holds a majority of
decision-making power and is “in charge”. Restrictiveness departs from an egalitarian relationship because one partner
assumes the right to intrude upon the other's behavior, even when that behavior does not directly involve the restrictive
partner, as when restrictive partners prohibit their partners from spending time with certain individuals. Disparagement occurs
when one partner fails to equally value the other partner and has an overall negative appraisal of his or her partner's worth.
4. Hypotheses
The empirical studies reviewed led to the following hypotheses:
1. The largest single category of PV is bidirectional violence, i.e. both partners engage in physical assault. The next
most frequently occurring pattern is “female-only”, i.e., the female partner is violent and the male partner is not. The
least frequently occurring pattern is “male-only”.
2. Dominance by one partner, regardless of whether it is the male or female partner, is associated with an increased
probability of violence.
5. Methods
5.1. The International Dating Violence Study
This research is part of the International Dating Violence Study, which is being conducted by a consortium of
researchers in all major world regions. Each consortium member used the same core questionnaire, except for the final
section, which was reserved for each member to add questions about issues of specific local or theoretical interest. A
detailed description of the study, including the questionnaire and all other key documents, is available on the website http://
pubpages.unh.edu/~mas2, and in previous articles reporting results from this study (Douglas & Straus, 2006; Straus, 2004;
Straus & International Dating Violence Research Consortium, 2004). A cross-national study may be the most appropriate
way to test the hypotheses because of the large nation-to-nation differences in the prevalence of violence against women
and male dominance (García-Moreno, Jansen, Ellsberg, Heise, & Watts, 2005; Krug et al., 2002).
5.2. Sample
This article presents results for a convenience sample of 4,239 students at 68 universities in 32 nations. The data
were usually obtained by administering a questionnaire during regularly scheduled classes. Most of the classes were in
psychology, sociology, criminology, and family studies. The percentage of students who chose to participate and
deposited a completed questionnaire ranged from 42 to 100%, with most participation rates ranging from 85 to 95%.
The results describe what was found for the students in those classes in each country and cannot be taken as representative
of students in general. Nevertheless, there is strong evidence that the behavior and opinions of the students in this study
reflect differences in the national contexts in which they are living (see Section 7.1 and Straus, 2007a).
Only students who were or had been in a relationship of a month or more were included in the analyses. The
questionnaires were examined for aberrant response patterns such as an implausibly high frequency of rare events, for
example, 10 instances of attacking a partner with a knife or gun in the past year; or inconsistent answers, for example,
reporting an injury but no assault. Based on this screening method, 6.2% of the respondents were dropped from the
sample. A total of 13,601 students completed all the questions used in the current analyses to be presented. The national
setting median sample size was 241 (range 90 to 4040).
5.3. Questionnaire administration
The data were gathered using procedures reviewed by and approved by the boards for protection of human subjects
at each of the universities in the study. The purpose of the study and the right to refuse to participate were explained to
all students. They were assured of anonymity and confidentiality, and given a debriefing form that explained the study
in more detail. They also provided contact information for area social service agencies should they need assistance.
5.4. Measure of dominance
Dominance by the partner who completed the questionnaire was measured by the Dominance scale of the Personal
and Relationships Profile (Straus, Hamby, Boney-McCoy, & Sugarman, 1999; Straus & Mouradian, 1999). This scale
is a nine item short form of the Dominance scale developed by Hamby (1996). Examples of the items are “I generally
have the final say when my partner and I disagree” and “My partner needs to remember that I am in charge”. The
response categories are 1 = Strongly Disagree, 2 = Disagree, 3 = Agree, and 4 = Strongly Agree. The scale score is the
mean of the nine items. The theoretical range of scores is from 1 to 4. The actual range for this sample was 1 to 3.96
(mean = 1.95, SD = 0.39).
5.5. Measures of partner violence
5.5.1. The CTS2

Physical assault was measured by the revised Conflict Tactics Scales or CTS2 (Straus, Hamby, Boney-McCoy, &
Sugarman, 1996). In the past 25 years, the CTS has been used in hundreds of studies, mostly in North America, but also in
many other countries. It has demonstrated cross-cultural reliability and validity (Archer, 1999; Straus, 1990a, 2004). This
research used the CTS2 scale for physical assault, both overall, and the subscale for severe assault. Most of the assaults
were in the “minor” category. Because severe violence is considered a unique phenomenon with a different etiology
(Johnson & Ferraro, 2000; Straus, 1990b), analyses were conducted for both overall PV, and separately for severe violence.
5.5.2. Mutuality types

This article also uses the recently published “Mutuality Types” (Straus & Douglas, 2004) which classifies
relationships where violence occurred into three mutually exclusive categories of perpetration: Male-Only, Female-
Only, and Both Violent. This typology enables research on an important aspect of PV that, despite its use in studies
such as the National Comorbidity Study (Kessler et al., 2001), is seldom investigated: the degree to which PV is mutual
(i.e., both partners are violent) or one-sided. The CTS makes this possible, even when only one partner has completed
the questionnaire, because CTS questions are asked in pairs. The first question in each pair asks about the behavior of
the study participant and the second question asks about the behavior of the partner. To control for possible gender bias
in reporting on both their own behavior and that of their partner, all analyses examining the links between dominance
and PV were done separately for male and female participants.
5.5.3. Physical assault

The CTS2 items to measure “minor” assault are: (1) pushed or shoved, (2) grabbed, (3) slapped, (4) threw something
at partner and (5) twisted arm or hair. The items in the “severe” assault scale are: (1) punched or hit a partner, (2) kicked,
(3) choked, (4) slammed against a wall, (5) beat up, (6) burned or scalded and (7) used a knife or gun on partner. The
“overall” rate of partner assault includes all of these items. Minor assault was scored as having occurred if one or more
of the behaviors had occurred in the past year. Severe assault was scored in the same way. Because most of the assaults
were minor, the minor assault scale and the overall or “any” assault scales largely overlap. To avoid this redundancy,
this article omits the minor assault variable and presents results for any assault and for severe assault. Separate results
are presented for mutuality of severe assaults, i.e. both severely assault.
5.6. Limited Disclosure Scale
In research on self-reported criminal behavior, differences between groups could reflect differences in willingness
to report socially undesirable behaviors as much or more than real differences in crime. To deal with this threat to
validity, we controlled for scores on a scale which measures the tendency to avoid reporting socially undesirable
behavior—the Limited Disclosure Scale of the Personal and Relationships Profile (Straus et al., 1999; Straus &
Mouradian, 1999). This is a13-item scale based on Reynolds short-form Social Desirability scale (Reynolds, 1982).
It asks about behaviors and emotions that are slightly undesirable but true of most people, such as “I sometimes try to
get even rather than forgive and forget”. The more items a respondent denies, the more likely a respondent will avoid
reporting PV. The response categories and the theoretical range of the Limited Disclosure Scale are the same as for
the Dominance scale (1 through 4). The scores for students in each national context ranged from 2.38 to 2.91; the
median score was 2.61.
5.7. Gender symmetry
Two aspects of gender symmetry will be investigated: symmetry in perpetration and symmetry in a key risk factor
for PV — dominance in the relationship. The criterion of symmetry in perpetration is that the percent of women who
physically assault a male partner is as high or higher than the percent of men who physically assault a female partner.
In addition, this must apply to both minor and severe assaults (as defined above). The criterion for symmetry in
dominance is that the mean Dominance scale scores of the women in this study are as high or higher than those of the
men, and that score on the Dominance scale is significantly related to perpetration of PV by women as well as by
men.
5.8. Demographic characteristics
5.8.1. Gender
Seventy one percent of the students were female because the questionnaires were administered in social science
courses that tend to have a heavy concentration of female enrollments. Because this study is focused on issues in which
gender differences are important, the analyses either controlled for gender or were replicated for male and female
students.
5.8.2. Age
Students' ages ranged from 18 to 40, with a median of 22.3. It is well established that younger ages are associated
with higher rates of violent crime, including PV (Stets & Straus, 1989).
5.8.3. Relationship length

The length of the relationships varied greatly. Whereas only 9.7% had been in their current relationship for one
month, 38% had been in the relationships from two to 12 months. Because relationships change over time, it was
important to control for the length of time the couple had been together.
6. Prevalence of partner violence
6.1. Assault rates
6.1.1. Overall assault rate

Table 1 gives the percent of students in each national setting who reported a physical attack on a dating partner
during the 12 months prior to completing the questionnaire. The data in the columns for violence by men are based on
the reports by male students, and the data on violence by women are based on the reports of female students.
Table 1
Prevalence of overall assault and severe assault, by national setting and gender, as reported by men and women (in rank order of male overall assault
rate)
National Setting Minor assault only Severe assault
Males Females F % of M Males Females F % of M
All (median)⁎⁎ 24.4 31.6 129 7.6 10.6 139
Iran 95.5 71.4 74 18.2 15.7 86
South Africa 42.9 39.3 91 42.9 14.0 32
Greece 38.5 25.5 66 17.6 13.9 79
India 35.5 31.0 87 7.4 13.8 186
Tanzania 31.7 43.8 138 14.8 25.7 173
Netherlands 31.3 31.8 101 8.3 4.5 54
Malta⁎ 30.4 15.8 52 .0 5.3 ?
Romania 29.2 32.2 110 8.7 10.8 124
Belgium 28.9 34.7 120 7.2 11.6 161
Mexico 26.7 47.3 177 13.3 17.2 129
Switzerland 26.7 24.2 90 6.9 5.4 78
United States 26.6 31.5 118 8.9 12.1 136
Hungary 26.5 21.0 79 12.2 10.5 86
Venezuela 25.3 23.7 93 14.5 9.6 66
Japan 24.6 18.3 74 6.6 10.2 154
Great Britain 24.6 39.7 161 7.0 15.6 222
South Korea 24.3 37.4 153 8.1 19.3 238
Germany 24.0 27.9 116 6.1 7.7 126
Russia 23.8 38.1 160 8.6 15.1 175
Hong Kong⁎⁎ 23.5 42.8 182 6.8 19.4 285
China⁎⁎ 22.4 42.3 188 8.8 20.3 230
Lithuania⁎⁎ 22.3 39.3 176 4.5 10.5 233
Brazil 21.8 23.3 106 6.5 5.7 87
Israel 21.4 18.0 84 8.9 7.2 80
Canada⁎ 19.3 27.1 140 6.6 9.5 143
Sweden 18.8 17.7 94 1.3 2.0 153
Australia 18.4 20.7 112 7.9 8.7 110
Taiwan 18.4 41.9 227 15.8 25.6 162
Guatemala 17.0 32.1 188 5.7 8.6 150
New Zealand 14.3 31.7 221 3.7 12.9 348
Portugal 14.2 17.8 125 5.3 5.0 94
Singapore 10.0 28.2 282 .0 7.0 ?
⁎ and ⁎⁎ indicate significant difference between male and female percentages at .05 and .01 level based on X2 for 2 by 3 cross tabulations (gender by
no violence, minor violence only) for each national setting.
? Indicates that the F % of M percent for Severe Violence could not be computed because the rate for males was 0. The effect is to underestimate the
median F % of M.
The first pair of columns in Table 1 under the heading Overall Assault gives the percent of male and female students
that physically assaulted a dating partner. The first row of these two columns shows that, on average for all the students
in the study, about one-quarter of both male and female students had physically attacked a partner during that year.
These are extremely high rates, but they are on the low side of the range of rates from a large number of mostly North
American studies which have found rates in the 20 to 40% range (Archer, 2000; Sugarman & Hotaling, 1989).
It is also important to note that there were very large differences between national settings in the percent who
assaulted a partner. The column for assaults by male students shows that the rates for any violence ranged from a high
of 95.5% in Iran to a low of 10% in Singapore.
6.1.2. Severe assault

Most of the attacks tabulated in the Overall Assault columns were “minor violence” such as slapping or throwing
things at a partner in anger. However, the columns headed Severe Assault in Table 1 show much lower, but still high,
rates of “severe violence”. These are attacks such as punching, kicking, and hitting with an object, which have a higher
probability of causing an injury. Overall, 8% of male students and 11% of female students in this study severely
attacked a dating partner. Again the rates obtained when students from all national contexts are combined are similar for
males and females, although very slightly higher for assaults by females. As in the case of the overall assault rate, there
are large differences between national contexts in the rate of severely assaulting a dating partner. The column for Males
in Table 1 shows that the rates ranged from a high of 42.9% in South Africa to zero in Malta and Singapore.
6.1.3. Ratio of male to female assaults

The column headed “F % of M” gives the percent that the rate of assault by females is of the rate of assault by males.
The row for all nations together shows that the female percent of the male rate is 129, i.e. that perpetration by the
women in this study is 29% greater than the perpetration rate of the men in this study. For severe violence, the
difference is somewhat greater—a 39% higher rate for women than for men. Examining the rows of Table 1 shows that
the pattern of higher rates of assault by women is found in about two thirds of the 32 national settings in this study.
Thus, as was also shown in a meta-analysis of previous studies (Archer, 2000), the predominant pattern is for a larger
percentage of women than men to physically attack their partner.
6.2. Bidirectionality and asymmetry
6.2.1. Prevalence of bidirectional and asymmetric violence

Fig. 1 is based on the sub-group of 4,239 students who reported one or more incidents of violence because they are
the only students for whom the question of bidirectionality is relevant. The left panel of Fig. 1 presents the results on
Fig. 1. Percent of violent couples in each mutuality type.

Table 2
Prevalence and bidirectionality of any physical violence
Nation Assault rate (%) Bidirectionality type a
Male-only % Female-only % Both Violent%
All (median) 31.2 9.9 21.4 68.6
Iran 77.1 4.0 1.3 94.6
Mexico 44.1 6.6 15.5 77.7
South Africa 39.5 7.6 10.2 82.0
Great Britain 37.5 4.2 19.2 76.5
Tanzania 37.4 6.9 1.7 91.3
Hong Kong⁎⁎ 37.0 7.7 36.9 55.2
Taiwan 36.1 6.7 25.4 67.8
China⁎⁎ 34.7 6.9 31.7 61.2
Lithuania⁎⁎ 33.9 5.6 23.5 70.7
Belgium⁎⁎ 33.4 11.2 19.8 68.8
India 32.3 8.3 16.6 75.0
South Korea 32.2 8.0 19.3 72.5
Russia 32.2 2.8 27.1 70.0
Romania 31.9 6.9 25.5 67.4
Netherlands⁎ 31.7 4.6 24.2 71.0
United States 30.0 9.7 20.6 69.6
Greece 28.5 26.2 16.2 57.5
New Zealand⁎ 27.9 8.7 28.2 63.0
Germany 26.6 12.4 24.8 62.7
Canada 24.9 9.4 22.1 68.3
Switzerland⁎ 24.7 9.4 25.8 64.7
Venezuela 24.3 17.4 19.0 63.4
Guatemala 24.2 8.3 25.0 66.6
Hungary 22.7 12.5 17.5 70.0
Singapore 22.7 9.4 32.0 58.4
Brazil 22.0 15.2 13.5 71.1
Japan 21.4 7.6 11.5 80.7
Australia 20.2 14.0 21.0 64.9
Malta 19.1 21.7 30.4 47.8
Israel⁎ 18.6 9.8 27.8 62.3
Sweden 17.9 11.8 28.4 59.7
Portugal⁎ 16.6 13.6 22.7 63.6
⁎ and ⁎⁎ indicate significant difference between male and female percentages at .05 and .01 level.
a
Percentages are based on the 4239 participants who reported a violent act by themselves or their partner. The n's are 420 Male-Only, 908 Female-
Only, and 2911 Bidirectional. The n's for each row can be computed using the n's in each row of Table 1 and the percentages in each mutuality type.
symmetry in PV using the Overall Assault measure. It shows that, regardless of whether the data were reported by male
or female respondents, in over two thirds of the cases, both partners were violent, Female-Only violence characterized a
fifth of the cases, and Male-Only violence was found in one out of ten couples. The right panel of Fig. 1 gives the
results using the Severe Assault measure. Again, the results based on reports by males and females are very similar, and
they show that the percent of bidirectional violence is still the largest category by far. The least frequently occurring
type is Male-Only.
6.2.2. National setting differences

Tables 2 and 3 compare national settings on bidirectionality types as reported by the combined samples of male and
female students. The national settings are arrayed in rank order of the percent of Male-Only violence. Table 2, which
uses the Overall Assault measure, shows large differences between the national settings. The percent in the Male-Only
category ranges from a high of 26.2% to a low of 2.8. There is no national setting in which “Male-Only” is the largest of
the three bidirectionality categories. Even in the two nations with the largest percent of Male-Only violence (Greece
and Malta) the percentages (26.2% and 21.7%) are far from even a simple majority. In every one of the 32 national
settings, bidirectional violence is the largest category.
Table 3
Prevalence and bidirectionality of severe physical violence
National Assault rate (%) Bidirectionality type
setting
Male-Only % Female-Only % Both Violent%
All (median) 10.8 15.7 29.4 54.8
Taiwan 23.2 2.7 37.8 59.4
Tanzania 19.8 21.2 12.1 66.6
Mexico 16.5 16.2 18.9 64.8
Iran 16.3 16.6 16.6 66.6
South Africa 16.1 18.7 25.0 56.2
China 15.9 19.7 41.3 38.8
Hong Kong 15.5 9.0 43.4 47.4
Greece 14.8 24.3 17.0 58.5
South Korea 14.8 6.4 22.5 70.9
Great Britain 14.4 11.2 28.1 60.5
Russia⁎ 12.4 7.5 30.1 62.2
India 11.9 15.3 23.0 61.5
Venezuela 11.4 21.8 15.6 62.5
Hungary 11.0 15.7 42.1 42.1
United States 11.0 15.2 28.1 56.6
New Zealand 10.9 0.0 60.0 40.0
Belgium 10.6 18.3 33.3 48.2
Romania 10.5 21.2 36.3 42.4
Canada 8.6 12.1 25.2 62.6
Australia 8.5 20.0 28.0 52.0
Lithuania⁎ 8.5 15.1 36.3 48.4
Japan 8.3 10.0 10.0 80.0
Israel 7.5 15.3 23.0 61.5
Germany 7.2 17.3 28.2 54.3
Guatemala 7.1 16.6 27.7 55.5
Brazil 5.9 11.7 17.6 70.5
Switzerland 5.7 9.5 28.5 61.9
Portugal 5.0 21.7 8.7 69.5
Netherlands 4.9 8.7 43.4 47.8
Singapore 4.9 7.6 46.1 46.1
Malta 4.0 37.5 50.0 12.5
Sweden 1.7 37.5 37.5 25.0
⁎ and ⁎⁎ indicate significant difference between male and female percentages at .05 and .01 level.
For severe Violence, the results in Table 3 parallel those for overall assaults. In none of the 32 National settings is
Male-Only the largest category. Moreover, for the two national settings that are highest in the percent of Male-Only
violence (Malta and Sweden), it is important to keep in mind that these are percentages of relationships in which
violence occurred. Malta and Sweden are also national settings with very low rates of severe violence. Thus the high
percentage of Male-Only violence represents a large piece of a small pie. In all 32 national settings, Both Violent is the
largest category of Severe Violence.
6.2.3. Age and partner violence

The results on bidirectionality for university student dating couples are similar in some respects to that found in
research on married couples, and different in other respects. The results are similar in that the most frequent pattern is
bidirectional violence. However, they differ in that previous studies of married couples have found about 50%
bidirectional compared to 69% in this study. In addition, with married couples, the Male-Only and Female-Only types
are about equally prevalent. The differences are probably the result of the age of the sample. PV, like other types of
violence, decreases rapidly with age, from the 30–40% rate typically found for all couples (married or dating) in the age
group of this sample, to about 12% for US couples at age 40 (the median age of US married couples), and continues to
slowly decline with increasing age (Stets & Straus, 1989; Suitor, Pillemer, & Straus, 1990).
6.2.4. Symmetry in severity

Although the results just presented show that symmetry in prevalence applies to severe as wall as minor assaults, the
analysis did not take into account whether one partner engaged in severe and the other in only minor assaults. This
possibility was investigated by cross tabulating the severity level of the violence the student perpetrated by the severity
of the violence of the partner. For students in the minor-only level category, the percent whose partners engaged in
severe assaults was 13.4% of the partners of men and 9.9% of the partners of women. For male students who engaged in
severe assault on a partner, the female partner restricted the attacks to the minor assaults in 14% of the cases, whereas
male partners of women restricted their violence to the minor level in 25% of the cases. Of students who perpetrated
severe assaults, 85% of the partners of men, and 74% of the partners of women, also engaged in severe assaults. Thus,
reciprocity in severity is predominant in bidirectional violence, but when the partner engages in more severe violence
than the student reported for him or herself, it was more often the female partners of men. Therefore, to the extent that
the previously presented results misrepresent the degree of symmetry because the category Both Violent does not take
into account differences in severity level, the misrepresentation is in failing to bring out that when both partners are
violent, more women than men tend to engage in violence that is more severe than that of their partner. In evaluating
these results, it is important to keep in mind that although there may be symmetry in perpetration, at both the minor and
severe levels, men inflict more physical injury than women.
Table 4
Dominance scale (in rank order of mean male Dominance score)
Nation Mean score of:
Males Females
ALL(median) 1.98 1.99
Tanzania 2.38 2.38
Russia⁎⁎ 2.37 2.21
Iran 2.27 2.32
Taiwan 2.23 2.28
China⁎⁎ 2.22 2.15
Greece 2.17 1.98
South Korea 2.17 2.25
Lithuania⁎⁎ 2.11 2.20
Hong Kong 2.10 2.16
India 2.10 2.18
Hungary⁎ 2.08 1.93
Mexico 2.03 2.10
Venezuela⁎ 2.03 1.93
South Africa 2.01 2.08
Romania 2.00 2.01
Guatemala 1.97 1.98
Brazil 1.96 1.94
Singapore 1.96 2.03
United States 1.93 1.91
Japan 1.90 1.97
Portugal 1.89 1.89
Germany 1.82 1.90
Israel 1.81 1.86
Australia 1.80 1.83
Great Britain 1.80 1.87
Belgium 1.77 1.80
Malta⁎⁎ 1.75 1.95
New Zealand 1.75 1.82
Canada⁎⁎ 1.73 1.81
Switzerland 1.73 1.76
Netherlands 1.67 1.70
Sweden 1.65 1.68
⁎ and ⁎⁎ indicate significant difference between mean scores for males and females at .05 and .01 level.
7. Dominance by males and females
The column headed Males in Table 4 gives the mean Dominance scale scores of the male students in each national
setting. The national settings are arrayed in rank order according to these scores. The nation with the highest score for
Dominance by male partners is Tanzania, which is also the least modernized of the 32 nations in this study. The four
national settings which are the next most male-dominant are Russia, Iran, Taiwan and mainland China, respectively.
The national setting in which male students have the lowest average Dominance score is Sweden, which is a nation that
has led the way in steps to promote gender equality. The other four of the five least male-dominant national settings are
Netherlands, Canada, Switzerland, and Malta.
The rank order of national settings discussed above is consistent with the idea that, among nation states, greater
economic development and modernity is associated with a waning of the traditional pattern of male dominance.
However, comparison of the Dominance scale scores of men and women is not consistent with the idea that men are
more dominant in couple relationships. The row for All Nations in Table 4 shows that for all students in the study, the
mean Dominance score of the women is almost the same as that for men. Overall, the Dominance scale scores are
higher for women than for men in 24 of the 32 nations, and in all 12 of the nations with the lowest scores for male
dominance. Although the differences are small, they are not consistent with the large body of evidence showing greater
male power in most societies (García-Moreno et al., 2005; Sugarman & Straus, 1988).
7.1. Validity of the dominance scale
The discrepancy between the nearly equal scores of male and female students on the Dominance scale raises
questions about the validity of this scale. A standard way of examining the validity of a measure is to determine the
degree to which it is correlated with another measure of known validity. This was done by correlating the Dominance
scale with scores for the United Nations Gender Empowerment Index (as given in the Human Development Report
2005, an independent report commissioned by the United Nations Development Programme [http://hdr.undp.org/]).
The Gender Empowerment scores were added to the data file for the 29 nations included in both this study and the UN
study. Partial correlation analysis, controlling for the mean score of students in each national setting on the Limited
Disclosure Scale, were computed. The partial correlation of − .69 indicates that the more Gender Empowerment, the
lower the Dominance score of the men in this study. For example, Tanzania has the lowest Gender Empowerment score
and also the highest Dominance score of the 29 national settings where both measures were available; and Sweden has
the highest Gender Empowerment score and the lowest Dominance scale score. Thus, the Dominance scale scores for
the men in this study are highly consistent with the widely used Gender Empowerment Measure.
8. Relation of dominance to partner violence
A previous article tested the idea that the etiology of PV by women is different than violence by men (Medeiros &
Straus, 2006). That study used all 23 risk factors measured by the Personal and Relationships Profile, but only for a
sample of University of New Hampshire students. There is insufficient space in this article to present that mass of data
for the 32 national settings. However, there is sufficient space to present the results for a risk factor that is central to the
feminist theory of PV — Dominance by one partner. According to feminist theory, dominance by the male partner
should be closely related to male assaults on female partners. Dominance by female partners is not discussed in feminist
analyses, so no inference can be made about its relation to violence against a male partner. However, the results of the six
studies cited earlier suggest that dominance by a female partner will also be related to an increased probability of PV.
The results of the multinomial logistic regression analysis to investigate the degree to which dominance is part of the
etiology of PV by women as well as by men are presented in Table 5 for violence in general, and in Table 6 for Severe
Violence. To help focus on the central issue, the rows for the Dominance scale and the odds ratio showing the relation
of Dominance to PV are in bold type. To further help understand the relationships between dominance and PV, the
regression coefficients were converted to estimated probabilities for each type of violence and those estimates are
graphed in Figs. 2–4.
Table 5 uses the dependent variable whether there had been any violence in the relationship during the previous
12 months. Most of the violent acts in that overall measure of violence are relatively minor, for example, slapping and
throwing things at a partner. Table 6 presents the results using severe violence such as punching, choking, and attacks
Table 5
Logistic regression of relation of dominance by one partner to violence bidirectionality types, by gender (all violence)
Dependent variable Independent variables B Standard Wald Significance Odds 95% confidence interval
error ratio
Lower bound Upper bound
A. Male respondents
Male- Only Violence Male dominance .830 .236 12.381 .000 2.293 .932 1.007
Age (years) − .032 .020 2.603 .107 .969 1.009 1.052
Relationship length (months) .030 .011 7.768 .005 1.030 .757 1.092
Socioeconomic Status Scale − .095 .093 1.043 .307 .909 .540 1.603
Limited Disclosure Scale − .072 .278 .068 .795 .930 .540 1.603
Female-Only violence Male dominance .676 .198 11.635 .001 1.965 .942 1.003
Age (years) − .028 .016 3.069 .080 .972 1.020 1.056
Socioeconomic Status Scale .053 .079 .453 .501 1.054 .495 1.227
Limited Disclosure Scale − .249 .232 1.158 .282 .779 .495 1.227
Both Violent Male dominance 1.131 .108 109.676 .000 3.098 .958 .991
Age (years) − .026 .009 9.362 .002 .974 1.039 1.059
Socioeconomic Status Scale .039 .043 .679 .410 1.036 .238 .399
Limited Disclosure Scale 1.177 .131 80.343 .000 .308 .238 .399
B. Female respondents
Male- Only Violence Female dominance .967 .170 32.409 .000 2.629 .980 1.016
Age (years) − .002 .009 .044 .833 .998 1.016 1.045
Limited Disclosure Scale − .104 .177 .345 .557 .901 1.885 3.668
Female- Only Violence Female dominance 1.231 .104 139.649 .000 3.425 .972 .996
Age (years) − .016 .006 6.681 .010 .984 1.025 1.044
Socioeconomic Status Scale − .016 .040 .156 .693 .984 .333 .518
Limited Disclosure Scale − .879 .113 60.369 .000 .415 .333 .518
Both Violent Female dominance 1.439 .074 382.723 .000 4.215 .954 .972
Age (years) − .038 .005 63.507 .000 .963 1.047 1.060
Socioeconomic Status Scale .001 .027 .001 .980 1.001 .302 .412
Limited Disclosure Scale − 1.042 .079 175.321 .000 .353 .302 .412
with an object. It is necessary to consider severe violence separately because it is often claimed that male predominance
in partner violence is particularly great for this level of violence, and that the etiology of severe violence is different for
severe violence.
8.1. Dominance and overall violence
8.1.1. Dominance by men and Male-Only violence

Part A of Table 5 refers to the behavior of the male students in this study, as reported by male students. The entry in
the column of odds ratios for Dominance in the panel for Male-Only violence shows an odds ratio of 2.29. This
indicates that each increase of one point on the four point Dominance scale increases the probability of violence by
male students 2.29 times. Of the other four variables in the Male-Only panel, only one—length of the relationship—is
significantly related to Male-Only violence. The odds ratio for age of 1.03 indicates that each additional month the
relationship has been ongoing increases the odds of Male-Only violence 1.03, or 3%. This may seem like a small odds
ratio but it is statistically significant and, for long-standing relationships such as one of ten months, it would mean a 3%
increase for each month, or a total of a 30% increase in the odds of violence compared to a just established relationship.
The age of the student, the socioeconomic status of the student's parents, and score on the Socially Desirable
Responding scale are not associated with an increase in the odds of Male-Only violence.
Table 6
Logistic regression of relation of dominance by one partner to violence bidirectionality types, by gender (severe violence)
Dependent variable Independent variables B Standard Wald Significance Odds 95% confidence interval
error ratio
Lower bound Upper bound
A. Male respondents
Male- Only Violence Male dominance 1.638 .322 25.860 .000 5.147 2.737 9.677
Age (years) −.017 .028 .363 .547 .983 .931 1.038
Relationship length (months) .038 .015 6.222 .013 1.038 1.008 1.070
Socioeconomic Status Scale −.044 .136 .106 .745 .957 .734 1.248
Limited Disclosure Scale −.411 .413 .988 .320 .663 .295 1.490
Female-Only violence Male dominance .517 .243 4.544 .033 1.678 1.043 2.700
Age (years) −.024 .020 1.400 .237 .977 .939 1.016
Limited Disclosure Scale −.634 .290 4.777 .029 .530 .300 .937
Both Violent Male dominance 1.503 .170 78.278 .000 4.497 3.223 6.274
Age (years) −.009 .014 .419 .517 .991 .964 1.019
B. Female respondents
Male-Only Violence Female dominance 1.611 .191 71.465 .000 5.010 3.448 7.280
Age (years) .005 .011 .226 .634 1.005 .984 1.027
Limited Disclosure Scale −.148 .212 .484 .487 .863 .569 1.308
Female-Only violence Female dominance 1.469 .139 112.396 .000 4.344 3.311 5.700
Age (years) −.024 .010 5.812 .016 .977 .958 .996
Socioeconomic Status Scale .111 .055 4.115 .042 1.117 1.004 1.243
Limited Disclosure Scale −1.064 .158 45.269 .000 .345 .253 .470
Both Violent Female dominance 1.742 .110 250.716 .000 5.708 4.601 7.082
Age (years) −.027 .008 12.349 .000 .973 .958 .988
8.1.2. Dominance by men and Female-Only violence

Moving down to the panel for Female-Only violence (as reported by male students), shows similar results.
Dominance by males in this study is associated with an average 1.96 times increase in the probability of Female-Only
violence for each one point increase in the Dominance scale. The only other significant relationship in the Female-Only
panel shows that the longer the relationship the higher the odds of Female-Only violence.
8.1.3. Dominance by men and bidirectional violence

The results in the Both Violent panel of Part A of Table 5 are similar to the results for Male-Only and Female-Only
violence, but there are three important differences. First, dominance by the male partner is associated with a three-fold
increase in the probability of both partners being violent. This is larger than the increase in the probability of Male-Only
or Female-Only violence. That is, dominance by a male partner is more strongly associated with bidirectional violence
than with Male-Only violence. Second, age was not related to the odds of Male-Only or Female-Only violence, but for
Both Violent, each additional year of age is associated with a small but statistically significant decrease in the odds of
both partners being violent. This is consistent with many studies showing that violent crime decreases with age. Finally,
the odds ratio of 0.31 for the Limited Disclosure Scale indicates that increase of one point on this four point scale
reduces the probability of both being violent by 69% (1.00 − 0.31 = .69 or 69%). In short, students identified by the
Limited Disclosure Scale as being unwilling to disclose minor types of socially undesirable behavior are also much less
likely to disclose a pattern of bidirectional violence in their dating relationships. This illustrates why it was necessary to
Fig. 2. Relationship between dominance scale score and probability of Male-Only violence, as reported by males and females.
Fig. 3. Relationship between dominance scale score and probability of Female-Only violence, as reported by males and females.
Fig. 4. Relationship between dominance scale score and probability of violence by both partners, as reported by males and females.
control for score on the Limited Disclosure Scale by including the score on the Limited Disclosure Scale as one of the
independent variables in all the models tested. Consequently, all the odds ratios in Tables 5 and 6 are adjusted to control
for score on the Limited Disclosure Scale.
8.1.4. Dominance by women and Male-Only violence

Part B of Table 5 gives the results for the behavior of female participants as reported by the female students. The
odds ratio of 2.629 in the Male-Only panel indicates that female dominance as reported by women is associated with
about a two and half times greater probability of the Male-Only pattern of PV. The only other significant odds ratios in
the Male-Only violence panel of Table 5B show that the probability of Male-Only violence increases by 3% for each
additional month the relationship has been in effect.
8.1.5. Dominance by women and Female-Only violence

The odds ratio in the middle panel of Table 5B shows that female dominance is much more strongly linked to
Female-Only violence than was shown for the relation of male dominance to Female-Only violence. That is, when
there is dominance by either partner, it increases the odds of Female-Only violence, but the increase is much greater for
female dominance. Three of the other four independent variables in the Dominance By Females panel of Table 5 are
also statistically significant.
8.1.6. Dominance by women and bidirectional violence

The odds ratios in the Both Violent panel of Part B of Table 5 follow a pattern that is similar to that for men, but the
effect of female dominance on the odds of bidirectional violence tends to be greater than the effect of male dominance. It
also shows that the age is related to a decrease in the odds of bidirectional violence, and that the longer the relationship
the greater the odds of bidirectional violence. Finally, as was found for male bidirectional violence as reported by males,
the higher the score on the Limited Disclosure Scale, the lower the odds of reporting bidirectional violence.
8.2. Predicted probability graphs
Figs. 1–3 provide a visual means for understanding the relationship between the Dominance scale and the
probability of PV, and also show the predicted probability of each of the three bidirectionality types. The upper line in
Fig. 2 plots the relationship between male dominance as reported by males and the estimated probability of Male-Only
assault. The estimated probabilities are after adjustment to control for the other three variables. The upper line shows
that the more dominant the male partner, the greater the probability that he had assaulted his partner during the previous
year. Similarly, the lower line plots the relationship between female dominance as reported by females and the
probability of Male-Only assault. It shows that female dominance is also associated with a greater chance of violence
by the male partner, but that the probability of Male-Only violence is not any greater at the highest level of female
dominance than at the middle level.
Fig. 3 shows that, although dominance by either the male or female partner is associated with an increased
probability of Female-Only violence, the relationship is weak for dominance by men (lower line) and strong for
dominance by women (upper line).
Fig. 4 shows that dominance by either the male or female partner is strongly associated with an increased probability
of both partners being violent. The probability of bidirectional violence increases from about 10% for both male and
female participants in this study who had the lowest Dominance scores, to over 50% for participants with the highest
scores.
8.3. Dominance and severe violence
Table 6 presents the results using the Severe Violence scale of the CTS as the dependent variable. As is the case in
the table using the overall violence scale (Table 5), Table 6 gives the results separately for dominance by the male
partner (Part A) and by the female partner (Part B).
The results for severe violence parallel the results for the overall violence scale but the odds ratios tend to be larger.
That is, dominance by either the male or female partner is more closely related to an increased probability of severe
violence than to minor violence, even though the overall rate of severe violence is much lower (10% versus 30%).
The top row of Table 6A shows that male dominance is associated with a five-fold increase in the odds of severe
Male-Only Violence. For female dominance, the top row of part B of Table 6 shows that female dominance is also
associated with a five-fold increase in severe Male-Only violence.
The middle panels of Table 6A and B show that male dominance is associated with a 1.67 times increase in the odds
of severe Female-Only violence, but the middle panel of Table 6B shows that female dominance is associated with a
much greater increase in the probability of Female-Only violence (a four fold increase). Finally, the lower panels of
Table 6A and B show that male dominance is associated with 4.5 times increase in the probability of severe
bidirectional violence, and the lower panel in Table 6B shows that female dominance is associated with an even greater
increase in the probability of severe bidirectional violence (a 5.7 fold increase). In general, dominance by either partner
is associated with an increased probability of severe violence, but dominance by the female partner increases the risk of
severe Female-Only and severe bidirectional violence, even more strongly than does dominance by the male partner.
9. Discussion
The results reported in this article are consistent with the first hypotheses—that bidirectional violence is the most
prevalent pattern, followed by Female-Only, and that Male-Only violence is the least frequently occurring pattern. The
results in the section on prevalence rates add cross-national evidence to the already overwhelming evidence from North
America which has found that about the same percentage of women are physically violent toward their partners as men,
and for young women, the percentage is higher than for men (Archer, 2000). In none of the 32 nations studied was
Male-Only violence the largest category. This applies to severe as well as minor violence. In a number of the national
contexts it was found in less than 10% of violent couples.
The results showing the predominance of bidirectional violence even in traditionally male-dominant societies may
seem implausible to many readers. However, they are consistent with results from the ongoing Global School-based
Health Survey conducted by the World Health Organization (WHO) among students 13 to 15 years old. The students
were asked if they had been hit, slapped or hurt on purpose by a boyfriend or girlfriend in the past 12 months. Results
for the first few countries show 15% of girls and 29% of boys in Jordan responded “yes”, as did 9% of girls and 16% of
boys in Namibia, 6% of girls and 8% of boys in Swaziland, and 18% of girls and 23% of boys in Zambia. In all five
countries, more girls hit partners than boys. Perhaps the results from both the International Dating Violence Study and
the WHO school study occurred because both studies are of dating relationships. The WHO could have answered that
question in another survey it conducted of married and cohabiting violence, but the organizers of that study followed
the usual practice of restricting the study to the victimization of women and refused requests to include questions on
perpetration by the women in the study.
Important as is the finding that the predominant pattern was one in which both partners were violent, the finding that
the second largest category was couples where the female partner was the only one to physically attack may be even
more important. Moreover, the predominance of bidirectional violence applies to severe violence, such as punching
and hitting with objects, as well as to minor violence. These results contradict the widely held belief that PV is
predominantly a crime committed by men. Indeed, almost every treatment and preventive effort is based on that
assumption, which these results suggest is false.
What explains the predominance of bidirectional violence? No doubt many factors contribute, as they do for other
aspects of PV. Self-defense is one of those factors, but as the studies reviewed earlier indicate, self-defense explains
only a small percentage of violence by either men or women. Another part of the explanation is suggested by the fact
that the second most common scenario is Female-Only. This may occur because, as Felson (2002) has shown, the
concept of chivalry did not go down with the Titanic. Consequently, up to a certain point, most men will not hit back if
struck by a woman because 'you don't hit women.' But if physical attacks by a female partner continue, sooner or later
male partners are likely to retaliate or follow her example when they have a grievance. To the extent that this scenario
occurs, it provides an additional part of the explanation for the predominance of bidirectional PV. The process by which
an escalation to violence takes place has been described for cases of male perpetration (Eisikovits, Winstok, & Gelles,
2002; Winstok, Eisikovits, & Gelles, 2002), and a parallel process may apply to violence by female partners.
The second hypothesis, that dominance by either partner, not just the male partner, is a risk factor for violence, was
also supported. In fact, this study found that, using data provided by women themselves, dominance by the female
partner is even more closely related to violence by women than is male dominance. The results on dominance as a risk
factor for violence, like the results on symmetry and asymmetry in perpetration, apply to both minor violence and
severe violence. This contradicts the belief that when women hit, the motives are different, and that male dominance is
the root cause of PV. Thus, the results for this hypothesis call into question another basic assumption of most prevention
and treatment programs. Unfortunately, bidirectional violence is the pattern of violence which is least likely to cease
and most likely to escalate (Feld & Straus, 1989). In addition, it is the type of PV that is most likely to result in physical
injury (Straus & Gozjolko, 2007; Whitaker et al., 2007), probably because it is the type with the greatest risk of
escalation.
9.1. Limitations
Although the results are clear and “strong” in the sense that the relationships are consistent and statistically
dependable, seem to apply world-wide, and are consistent with other studies, there are also limitations to keep in mind
when considering the implications that follow from the results.
9.1.1. Student sample

This study is based on a sample of university students rather than a sample of the general population, and it is a
convenience sample rather than a probability sample of students. Even if it were a representative sample, they would
not be representative of the nation because of their age and because university students tend to come from high
education families. Therefore, the results apply to this sample and it remains unknown whether they also apply more
broadly. The descriptive statistics are the results most likely to differ from those of the general population. For example,
a representative sample of married couples is unlikely to have much lower rates of PV than those reported because their
average age would be much older (about 40 rather than 20). The high rates of dating PV found by many studies stem
from their youthfulness (Stets & Straus, 1989). On the other hand, there is voluminous research on risk factors for
dating PV. With rare exception, the results are parallel to those found for general population samples. Thus the results
on the relation of Dominance to PV may have more general applicability than the students who participated in this
study.
This possible generality of the results is also indicated by the fact that, as reported in the Methods section, the
Dominance scale scores for each nation are highly correlated with the Gender Empowerment Index for each nation.
Other examples of congruence between nation-to-nation differences as measured in the International Dating Violence
Study and as measured by data from other sources are in Straus (2007a). These correlations suggest that, although the
students in this study are not representative samples, the differences between nations in student behavior reflect the
national context in which they are living.
9.1.2. Clinical versus community sample

A related sample issue is that a sample of a general population (such as a community or student survey) can differ
importantly from a “clinical” sample of men and women arrested for PV or men and women who seek assistance
because of PV. This difference is both a strength and a limitation. It is a strength because research with non-clinical
population samples is needed to guide prevention efforts. What is true of a clinical population often does not
necessarily apply to the general population (i.e., the target of primary prevention). To assume that it does apply has been
called the “clinical fallacy”. Conversely, what is true of the general population does not necessarily apply to clinical
populations. To assume that it does has been called the “representative sample fallacy” (Straus, 1990b). A simple
example, but one which has important implications for treatment, is the widely held belief that once PV starts, it may
escalate, but is not likely to cease without a strong intervention. That is correct for clients of shelters for battered
women. They would not be there if it had ceased. On the other hand, general population studies such as in Feld and
Straus (1989) consistently find high rates of cessation. Thus, policies and practices based on a clinical sample may not
apply to the general population, just as advice based on the general population may not apply to clinical populations.
Similarly, the results of this study concerning gender symmetry in perpetration and in etiology may not apply to
severely assaulted and oppressed women, such as those who seek help from a shelter for battered women, or to women
who are part of the small percent of violent couples (less than 1%) who have had violence progress to the point of police
intervention (Kaufman Kantor, & Straus, 1990). However, the five studies of clinical populations reviewed earlier
show that bidirectional violence also characterizes “clinical level” cases. The most important difference between
general population samples and clinical samples may not be in bidirectionality, but in the high rates of social and
psychological problems such as alcoholism, antisocial personality, and prior non-family crime (Ehrensaft et al., 2004;
Straus & Scott, in press) that characterize one or both partners. The title and content of a recent article epitomizes the
current situation: “Treatment for partner abuse: Time for a paradigm shift” (Stuart, 2005).
9.1.3. Cross-sectional data

Caution is also needed because the results are based on cross sectional data and may not reflect a cause-effect
relationship between dominance and PV. However, the analyses controlled for a number of variables that could produce
spurious results, such as confounding with socioeconomic status, and differences in willingness to report socially
undesirable behavior.
9.1.4. Self-defense
Another limitation of the study is that there is no direct evidence which contradicts the belief that PV by women is
primarily an act of self-defense. However, self-defense is unlikely to apply to the roughly one-quarter of cases where
the only violence was perpetrated by the female partner. Self-defense could, of course, apply to the two thirds of cases
where the violence was bidirectional. However, the results of this study, and those reviewed earlier, have found that
dominance by women is associated with bidirectional violence. This makes the self-defense explanation less plausible.
In addition, the studies that directly investigated self-defense (reviewed in the introduction) find that, for women as
well as for men, violence in self-defense applies to only a minority of cases of PV.
9.2. Theoretical and methodological implications
9.2.1. Theoretical implications

The results showing that bidirectional violence is the most prevalent form of PV in this sample; that Male-Only
violence is the least prevalent form; and the results showing that dominance by either the male or the female partner
(rather than just male dominance) increases the probability of PV—all call for a basic reorientation of the way PV is
conceptualized, especially because these results are consistent with results from other studies, including nationally
representative samples of the general population and studies of clinical level cases. It is the injustices and power
struggles that are associated with inequality, and the psychological problems of the partners that give rise to violence,
not just inequality in the form of male dominance. This study and other studies have found that male dominance in
couple relationships is not more prevalent than female dominance. This is not to deny the existence of male-dominant
oppressive relationships—or at the societal level, the fact that some societies are extremely male-dominant. But at the
individual couple level, such relationships are not typical among the students in the 32 nations studied.
9.2.2. Risk factors versus one-to-one causes

It is important for both theoretical understanding of domestic violence and clinical practice to keep in mind that
dominance is a “risk factor”, not a one-to-one cause for PV. For example, the probability of the participants in this study
assaulting a partner went from about 10% for those with the lowest Dominance scale score to about 50% for those with
the highest Dominance scores, or a five-fold increase. However, these same results also indicate that half of those with
the highest Dominance score did not assault their partner. Similarly, in the National Family Violence Survey, extremely
male-dominant partners had roughly ten times greater probability of assaulting a partner than did equalitarian men, but
that raised the rate from 2% to 20%, which means that 80% of extremely male-dominant partners did not assault their
partner in the year covered by this study. These same principles apply to all risk factors such as binge-drinking
(Kaufman Kantor & Straus, 1987) and corporal punishment as a child (Straus & Yodanis, 1996). Binge-drinkers and
those who were spanked a lot have much higher probability of assaulting a partner, but most do not physically assault
their partner.
9.2.3. Types of violence

Over the last 20 years, the fact that there are major differences within the category of “violent” has gained increasing
acceptance in principle. A simple but important difference is variation in the chronicity and severity of PV. The typical
pattern is an occasional episode of minor violence such as slapping and throwing objects at a partner, with only rare
injury. This is what Straus and colleagues called “ordinary” or “normal” (in the statistical sense) violence (Straus,
1990b; Straus et al., 1980, 2006)), and Johnson (1995) called “common couple” violence. These cases are likely to
differ in many ways from cases of chronic and severe assaults with a higher probability of injury—a pattern that
characterizes the experiences of many women who use services for female victims. However, those cases are only a
small fraction of cases revealed by community samples because such cases are rare in the first place, and because some
of the victims and perpetrators do not disclose what has happened to survey interviewers (Straus, 1990b).
More comprehensive typologies have been developed by Holtzworth-Munroe (Holtzworth-Munroe, Meehan,
Herron, & Stuart, 1999; Holtzworth-Munroe & Stuart, 1994) who distinguishes between Family Only, Borderline/
Dysphoric, and Generally Violent/Antisocial types; and by Johnson (1995), who identifies what he originally
identified as “patriarchal terrorism” and has renamed as “terroristic” violence (Johnson & Ferraro, 2000). Still another
aspect are the simple but crucial differences between Male-Only, Female-Only, and Both Violent types used for this
article.
The varying patterns of PV identified by these and other typologies, although increasingly recognized by both
researchers and clinicians, have only rarely been put into the actual practice of research or prevention and treatment.
The implication for research is that investigations of etiology and consequences must avoid grouping all cases together,
and instead examine the etiology and the consequence of each type. The same implication applies to prevention and
treatment.
9.3. Policy and practice implications
This study and those cited in the introduction revealed an overwhelming body of evidence that bidirectional
violence is the predominant pattern of perpetration; and this study, along with evidence from many other studies
(Medeiros & Straus, 2007), indicates that the etiology of PV is mostly parallel for men and women. The fact that
dominance by women as well as men was found to be a risk factor for violence is crucial for effective prevention and
treatment. Male dominance does need to be addressed, but so does female dominance, and many other family system
problems. In short, PV is more a gender-inclusive family system problem than it is a problem of a patriarchal social
system which enforces male dominance by violence. The fact that these results apply even in nations with an
established and pervasive patriarchal system is even a stronger indication of the need for prevention and treatment
programs to go beyond the feminist conceptualization of PV.
Unfortunately, going beyond the feminist model will be difficult because the organization, funding and staffing of
current prevention and treatment efforts are wedded to the patriarchal dominance theory as embodied, for example, in
the “Duluth” treatment program (Rosenbaum & Price, 2007), and use many methods to block deviation from that
model (Straus, 2006). If researchers or service providers do not declare allegiance to these articles of faith, they risk
being denied funding and ostracized, as documented in Straus (1990c, in press) and Holtzworth-Munroe (2005). A
recent example is “request for proposals” on research on physical and sexual violence against partners issued in
December 2005 by the National Institute of Justice. Both specified that applications which dealt with male victims
would not be considered for funding. A set of nine articles that provide the most comprehensive available review of risk
factors for family violence (Heyman & Slep, 2001) included an article on risk factors for male violence but nothing on
violence by women. This omission was in response to the interest expressed by the funding agency. Treatments with
individuals or couples (e.g., anger control programs, couple treatment) are specifically excluded from state standards
for batterer intervention programs in 43% of US states.
As stated earlier, the refusal to recognize the family system nature of PV and the multiple causes of PV have
hampered efforts to end domestic violence. This has resulted in deliberately ignoring half the perpetrators. The
domestic violence service system, including services for female victims, needs to replace the default-assumption
that PV is primarily the product of male dominance. Instead, the results from this study and those reviewed indicate
that the default-assumption needs to be that PV is predominantly bidirectional both in respect to physical assaults
and in respect to other kinds of bidirectional mistreatment by a partner; and that the risk factors are mostly the same
for males and females. Despite these obstacles, the situation is slowly changing, as evidenced by the two special
issues of this journal and by three recent books, all of which go beyond the “men defending patriarchy” conception
of PV (Dutton, 2006; Hamel & Nicholls, 2007; Whitaker et al., 2007). At the same time, service providers need to
remain alert to cases that do not fit the typical pattern, including cases which fit the classical image of an oppressed
and battered spouse. Although there are men who fall into this category, it is more often women. In addition, the
harmful effects of all levels of violence are greater for women, physically, psychologically, and economically.
Consequently, although services for male victims are needed, the need for services for female victims will continue
to be greater.
A number of changes in prevention and offender treatment programs follow from recognizing the predominance of
bidirectional PV, the family system nature of PV, and the parallel etiology of PV by male and female partners. These
include:
Replace the assumption that almost all PV is male-only, with the assumption that it could be bidirectional violence or
female-only, and that this needs to be determined at the very onset of remedial efforts.
Replace the single causal factor “patriarchal system” model with a multi-causal model which recognizes the
prevalence of psychological and social problems of both partners.
Replace male dominance as the major risk factor in need of change with dominance by either party, but only as one
of many risk factors that need attention.
Give equal attention to developing prevention programs targeted to violence by women and girls as well as to men
and boys.
Secondary prevention efforts need to be open to a variety of new approaches, of which one of the most promising is
restorative justice (Mills, 2003, 2006; Strang & Braithwaite, 2002)
Two recent books and a forthcoming book provide more specific implementation principles and guidelines for
putting into practice suggestions such as those listed above (Dutton, 2006; Hamel & Nicholls, 2007; Lutzker &
Whitaker, in press; Potter-Efron, 2007).
These changes in policy and practice, rather than weakening efforts to protect women, will enhance the protection of
women because violence by women is a major factor contributing to the victimization of women. When women are
violent, they are the partners most likely to be injured (Straus, 2005a,b). Therefore, efforts to end PV by women will
contribute to protecting women. It is time to make the prevention and treatment effort one that is aimed at ending all
family violence, starting with spanking children (Straus & Yodanis, 1996), not just violence against women. Only then
will women, as well as all other human beings, be safe in their own homes.
References
Anderson, K. L. (2002). Perpetrator or victim? Relationships between intimate partner violence and well-being. Journal of Marriage and the Family,
64(4), 851−863.
Archer, J. (1999). Assessment of the reliability of the Conflict Tactics Scales: A meta-analytic review. Journal of Interpersonal Violence, 14(12),
1263−1289.
Archer, J. (2000). Sex differences in aggression between heterosexual partners: A meta-analytic review. Psychological Bulletin, 126(5), 651−680.
Capaldi, D. M., & Owen, L. D. (2001). Physical aggression in a community sample of at-risk young couples: Gender comparisons for high frequency,
injury, and fear. Journal of Family Psychology, 15(3), 425−440.
Carrado, M., George, M. J., Loxam, E., Jones, L., & Templar, D. (1996). Aggression in British heterosexual relationships: A descriptive analysis.
Aggressive Behavior, 22, 401−415.
Cascardi, M., & Vivian, D. (1995). Context for specific episodes of marital violence: Gender and severity of violence differences. Journal of Family
Violence, 10(3), 265−293.
DeKeseredy, W. S., Saunders, D. G., Schwartz, M. D., & Shahid, A. (1997). The meanings and motives for women's use of violence in Canadian
college dating relationships: Results from a National Survey. Sociological Spectrum, 17, 199−222.
Douglas, E. M., & Straus, M. A. (2006). Assault and injury of dating partners by university students in 19 countries and its relation to corporal
punishment experienced as a child. European Journal of Criminology, 3, 293−318.
Dutton, D. G. (2006). Rethinking domestic violence. Vancouver, BC: University of British Columbia Press.
Ehrensaft, M. K., Moffitt, T. E., & Caspi, A. (2004). Clinically abusive relationships in an unselected birth cohort: Men's and women's participation
and developmental antecedents. Journal of Abnormal Psychology, 113(2), 258−271.
Ehrensaft, M. K., & Vivian, D. (1999). Is partner aggression related to appraisals of coercive control by a partner. Journal of Family Violence, 14(3),
251−266.
Eisikovits, Z., Winstok, Z., & Gelles, R. J. (2002). Structure and dynamics of escalation from the victim's perspective. Families in Society, 83(2),
142−152.
Feld, S. L., & Straus, M. A. (1989). Escalation and desistance of wife assault in marriage. Criminology, 27(1), 141−161.
Felson, R. B. (2002). Violence and gender reexamined. Washington, DC: American Psychological Press.
Felson, R. B., & Messner, S. F. (1998). Disentangling the effects of gender and intimacy on victim precipitation in homicide. Criminology, 36(2),
405−423.
Felson, R. B., & Outlaw, M. C. (2007). The control motive and marital violence. Violence and Victims, 22, 387−407.
Fiebert, M. S., & Gonzalez, D. M. (1997). College women who initiate assaults on their male partners and the reasons offered for such behavior.
Psychological Reports, 80, 583−590.
Follingstad, D. R., Wright, S., Lloyd, S., & Sebastian, J. A. (1991). Sex differences in motivations and effects in dating violence. Family Relations,
40(1), 51−57.
García-Moreno, C., Jansen, H. A. F. M., Ellsberg, M., Heise, L., & Watts, C. (2005). W.H.O. Multi-country study on women's health and domestic
violence against women: initial results on prevalence, health outcomes and women's responses. Geneva: World Health Organization.
Gelles, R., & Straus, M. (1988). Intimate violence: the causes and consequences of abuse in the American family. New York, NY: Simon & Schuster.
Giles-Sims, J. (1983). Wife battering: a systems theory approach. New York: Guilford Press.
Gondolf, E. W. (2002). Batterer intervention systems: issues, outcomes, and recommendations. Thousand Oaks: Sage Publications.
Hamberger, L. K., & Guse, C. E. (2002). Men's and women's use of intimate partner violence in clinical samples. Violence Against Women, 8(11),
1301−1331.
Hamby, S. L. (1996). The Dominance Scale: Preliminary psychometric properties. Violence and Victims, 11(3), 199−212.
Hamel, J. & Nicholls, T. (Eds.). (2007). Family interventions in domestic violence: a handbook of gender-inclusive theory and treatment. NY:
Springer.
Harned, M. S. (2001). Abused women or abused men? An examination of the context and outcomes of dating violence. Violence and Victims, 16(3),
269−285.
Heyman, R. E., & Slep, A. M. S. (2001). Risk factors for family violence: Introduction to the special series. Aggression and Violent Behavior, 6, 115−119.
Holtzworth-Munroe, A. (2005). Female perpetration of physical aggression against an intimate partner: A controversial new topic of study. Violence
and Victims, 20(2), 251−259.
Holtzworth-Munroe, A., Meehan, J. C., Herron, K., & Stuart, G. L. (1999). A typology of male batterers: An initial examination. In X. B. Arriaga &
S. Oskamp (Eds.), Violence in intimate relationships. Thousand Oaks: Sage Publications, Inc.
Holtzworth-Munroe, A., & Stuart, G. L. (1994). Typologies of male batterers: Three subtypes and the differences among them. Psychological
Bulletin, 116(3), 476−497.
Johnson, M. P. (1995). Patriarchal terrorism and common couple violence: Two forms of violence against women. Journal of Marriage and the
Family, 57(May), 283−294.
Johnson, M. P., & Ferraro, K. J. (2000). Research on domestic violence in the 1990s: Making distinctions. Journal of Marriage and the Family, 62
(4), 948−963.
Jurik, N. C., & Gregware, P. (1989). A method for murder: an interactionist analysis of homicides by women. Tempe, AZ: School of Justice Studies,
Arizona State University.
Kaufman Kantor, G., & Straus, M. A. (1987). The drunken bum theory of wife beating. Social Problems, 34, 213−230. In M. A. Straus & R. J. Gelles
(Eds.), Physical violence in American families: risk factors and adaptations to violence in 8145 families (pp. 213−230). New Brunswick, NJ:
Transaction Publishers.
Kaufman Kantor, G., & Straus, M. A. (1990). Response of victims and the police to assaults on wives. In M. A. Straus & R. J. Gelles (Eds.), Physical
violence in American families: risk factors and adaptations to violence in 8145 families (pp. 473−487). New Brunswick, NJ: Transaction
Publishers.
Kessler, R. C., Molnar, B. E., Feurer, I. D., & Appelbaum, M. (2001). Patterns and mental health predictors of domestic violence in the United States:
Results from the National Comorbidity Survey. International Journal of Law and Psychiatry, 24(4–5), 487−508.
Kim, J. -Y., & Emery, C. (2003). Marital power, conflict, norm consensus, and marital violence in a nationally representative sample of Korean
couples. Journal of Interpersonal Violence, 18(2), 197−219.
Krug, E. G., Dahlberg, L. L., Mercy, J. A., Zwi, A. B., & Lozano, R. (2002). World report on violence and health. Geneva: World Health Organization.
Laroche, D. (2005). Aspects of the context and consequences of domestic violence — Situational couple violence and intimate terrorism in Canada in
1999. Quebec: Government of Quebec: Institut de la statistique du Québec.
Lutzker, J. R. & Whitaker, D. J. (Eds.). (in press). Prevention of partner violence. Washington, DC: American Psychological Association.
McCarroll, J. E., Ursano, R. J., Fan, Z., & Newby, J. H. (2004). Patterns of mutual and nonmutual spouse abuse in the U.S. Army (1998–2002).
Violence and Victims, 19(4), 453−468.
Medeiros, R. A., & Straus, M. A. (2006). Risk factors for physical violence between dating partners: Implications for gender-inclusive prevention and
treatment of family violence. In J. C. Hamel & T. Nicholls (Eds.), Family approaches to domestic violence: a practioner's guide to gender-
inclusive research and treatment (pp. 59−87). NY: Springer also available at http://pubpages.unh.edu/~mas2
Medeiros, R. A., & Straus, M. A. (2007). Risk factors for physical violence between dating partners: Implications for gender-inclusive prevention and
treatment of family violence. In J. Hamel & T. Nicholls (Eds.), Family Interventions in Domestic Violence. NY: Springer.
Mills, L. G. (2003). Insult to injury: rethinking our responses to intimate abuse. Princeton, NJ: Princeton University Press.
Mills, L. G. (2006). The justice of recovery: How the state can heal the violence of crime. Hastings Law Journal, 57(3), 457−508.
Moffitt, T. E., Caspi, A., Rutter, M., & Silva, P. A. (2001). Sex differences in antisocial behavior. Cambridge, UK: Cambridge University Press.
Oswald, D. L., & Russell, B. L. (2006). Perceptions of sexual coercion in heterosexual dating relationships: The role of aggressor gender and tactics.
Journal of Sex Research, 43(1), 87−95.
Pearson, P. (1997). When she was bad: women and the myth of innocence. Toronto: Random House.
Potter-Efron, R. T. (2007). Anger, aggression, domestic violence, and substance abuse. In J. Hamel & T. Nicholls (Eds.), Family interventions in
domestic violence. NY: Springer.
Reynolds, W. M. (1982). Development of reliable and valid short forms of the Marlowe-Crowne Social Desirability Scale. Journal of Clinical
Psychology, 38, 119−125.
Rosenbaum, A., & Price, B. (2007). National Survey of Perpetrator Intervention Programs. Paper presented at the International Family Violence
Research Conference, Portsmouth, New Hampshire.
Sarantakos, S. (1998). Husband abuse as self-defense. Paper presented at the International Congress of Sociology.
Saunders, D. G. (1986). When battered women use violence: Husband-abuse or self-defense? Violence and Victims, 1(1), 47−60.
Sommer, R. (1996). Male and female perpetrated partner abuse: testing a diathesis-stress model. Winnipeg, Manitoba: University of Manitoba, PhD
Dissertation.
Stets, J. E. (1991). Psychological aggression in dating relationships: The role of interpersonal control. Journal of Family Violence, 6(1), 97−114.
Stets, J. E., & Hammons, S. A. (2002). Gender, control, and marital commitment. Journal of Family Issues, 23(1), 3−25.
Stets, J. E., & Pirog-Good, M. A. (1990). Interpersonal control and courtship aggression. Journal of Social and Personal Relationships, 7, 371−394.
Stets, J. E., & Straus, M. A. (1989). The marriage license as a hitting license: A comparison of assaults in dating, cohabiting, and married couples.
Journal of Family Violence, 4(2), 161−180 also reprinted in Straus and Gelles, 1990.
Strang, H., & Braithwaite, J. (2002). Restorative justice and family violence. New York: Cambridge University Press.
Straus, M. A. (1976). Sexual inequality, cultural norms, and wife-beating. In E. C. Viano (Ed.), Victims and Society (pp. 543−559). Washington, DC:
Visage Press.
Straus, M. A. (1990a). The Conflict Tactics Scales and its critics: An evaluation and new data on validity and reliability. In M. A. Straus & R. J. Gelles
(Eds.), Physical violence in American families: risk factors and adaptations to violence in 8145 families (pp. 49−73). New Brunswick, NJ:
Transaction Publications.
Straus, M. A. (1990b). Injury, frequency, and the representative sample fallacy in measuring wife beating and child abuse. In M. A. Straus & R. J.
Gelles (Eds.), Physical violence in American families: risk factors and adaptations to violence in 8145 families (pp. 75−89). New Brunswick, NJ:
Transaction Publications.
Straus, M. A. (1990c). The National Family Violence Surveys. In M. A. Straus & R. J. Gelles (Eds.), Physical violence in American families: risk
factors and adaptations to violence in 8145 families (pp. 3−16). New Brunswick: Transaction Publishers.
Straus, M. A. (2004). Cross-cultural reliability and validity of the Revised Conflict Tactics Scales: A study of university student dating couples in 17
nations. Cross-Cultural Research, 38, 407−432.
Straus, M.A. (2005a, July 10 – 13). Gender and partner violence in world perspective: some results from the International Dating Violence Study.
Paper presented at the 9th International Family Violence Research Conference, Portsmouth, NH.
Straus, M. A. (2005b). Women's violence toward men is a serious social problem. In D. R. Loseke R. J. Gelles & M. M. Cavanaugh (Eds.), Current
controversies on family violence (pp. 55−77). 2nd ed. Newbury Park: Sage Publications.
Straus, M. A. (2006). Future research on gender symmetry in physical assaults on partners. Violence Against Women, 12(11), 1086−1097.
Straus, M. A. (2007a). Validity of the International Dating Violence Study data. Violence against dating partners in world perspective: the
International Dating Violence Study. Durham, NH: Family Research Laboratory, University of New Hampshire.
Straus, M. A. (2007b). Why, despite overwhelming evidence, partner violence by women has not been perceived and often denied. Durham, NH:
Family Research Laboratory, University of New Hampshire.
Straus, M.A. (in press). Processes explaining the concealment and distortion of evidence on gender symmetry in partner violence. European Journal
of Criminal Policy and Research.
Straus, M. A., & Douglas, E. M. (2004). A short form of the revised Conflict Tactics Scales, and typologies for seventy and mutuality. Violence and
Victims, 19(5), 507−520.
Straus, M.A., Gelles, R.J., & Steinmetz, S.K. (1980 (2006)). Behind closed doors: violence in the American family. New York: Doubleday/Anchor
Books (Re-issued Transaction Publications, 2006 with a new forward by Richard J. Gelles and Murray A. Straus).
Straus, M.A., & Gozjolko, K. (2007). Intimate terrorism and injury of dating partners by male and female university students in 32 nations. Paper
presented at the Stockholm Criminology Prize Symposium.
Straus, M. A., Hamby, S. L., Boney-McCoy, S., & Sugarman, D. B. (1996). The revised Conflict Tactics Scales (CTS2): development and preliminary
psychometric data. Journal of Family Issues, 17(3), 283−316.
Straus, M. A., Hamby, S. L., Boney-McCoy, S., & Sugarman, D. (1999). The personal and relationships profile (PRP). Durham, NH: University of
New Hampshire, Family Research Laboratory. Available in: http://pubpages.unh.edu/~mas2/
Straus, M. A., International Dating Violence Research Consortium. (2004). Prevalence of violence against dating partners by male and female
university students worldwide. Violence Against Women, 10(7), 790−811.
Straus, M. A., & Members of the International Dating Violence Research Consortium. (2006). Unpublished data from The International Dating
Violence Study. Durham, New Hampshire: Research Laboratory, University of New Hampshire.
Straus, M. A., & Mouradian, V. E. (1999, November 19, 1999). Preliminary psychometric data for the Personal and Relationships Profile (PRP): a
multi-scale tool for clinical screening and research on partner violence. Paper presented at the American Society of Criminology, Toronto,
Ontario.
Straus, M. A., & Scott, K. (in press). Gender symmetry in partner violence: the evidence and the implications for primary prevention and treatment. In
J. R. Lutzker & D. J. Whitaker (Eds.), Prevention Of Partner Violence. Washington, DC: American Psychological Association.
Straus, M. A., & Yodanis, C. L. (1996). Corporal punishment in adolescence and physical assaults on spouses later in life: what accounts for
the link? Journal of Marriage and the Family, 58(4), 825−841. Also in Straus, M. A., & Medeiros, R. A. (in press). The primordial
violence: corporal punishment by parents, cognitive development, and crime. Physical Assaults On Spouses. Walnut Creek, CA: AltaMira
Press.
Stuart, R. B. (2005). Treatment for partner abuse: Time for a paradigm shift. Professional Psychology-Research and Practice, 36(3), 254−263.
Sugarman, D. B., & Hotaling, G. T. (1989). Dating violence: Prevalence, context, and risk markers. In A. A. Pirog-Good & J. E. Stets (Eds.), Violence
in dating relationships: Emerging social issues (pp. 3−31). New York: Praeger.
Sugarman, D. B., & Straus, M. A. (1988). Indicators of gender equality for American-states and regions. Social Indicators Research, 20(3), 229−270.
Sugihara, Y., & Warner, J. A. (2002). Dominance and domestic abuse among Mexican Americans: Gender differences in the etiology of violence in
intimate relationships. Journal of Family Violence, 17, 315−339.
Suitor, J. J., Pillemer, K., & Straus, M. A. (1990). Marital violence in a life course perspective. In M. A. Straus & R. J. Gelles (Eds.), Physical violence
in American families: risk factors and adaptations to violence in 8145 families (pp. 305−319). New Brunswick, NJ: Transaction.
Tang, S. -k. T. C. (1999). Marital power and aggression in a community sample of Hong Kong Chinese families. Journal of Interpersonal Violence,
14(6), 586−602.
United Nations Development Programme. (2005). Human Development Report 2005. http://hdr.undp.org/
Walker, L. E. (1984). The battered woman syndrome. New York: Springer.
Whitaker, D. J., Haileyesus, T., Swahn, M., & Saltzman, L. S. (2007). Differences in frequency of violence and reported injury between relationships
with reciprocal and nonreciprocal intimate partner violence. American Journal of Public Health, 97(5), 941−947.
Williams, S. L., & Frieze, I. H. (2005). Patterns of violent relationships, psychological distress, and marital satisfaction in a national sample of men
and women. Sex Roles, 52(11–12), 771−784.
Winstok, Z. (2007). Toward an interactional perspective on intimate partner violence. Aggression and Violent Behavior, 12(3), 348−363.
Winstok, Z., Eisikovits, Z., & Gelles, R. J. (2002). Structure and dynamics of escalation from the batterer's perspective. Families in Society, 83(2),
129−141.

Murray Straus

Загружено:

Сведения о документе

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Murray Straus

Загружено:

Авторское право:

Доступные форматы

Available online at www.sciencedirect.

Children and Youth Services Review 30 (2008) 252 – 275

Dominance and symmetry in partner violence by male and

Keywords: Domestic violence; Gender; Prevention; Assault; Patriarchy

⁎ Tel.: +1 603 862 2594.

2. Previous evidence on bidirectional perpetration

3. Research on gender symmetry in dominance

3.1. Symmetry in dominance

3.2. Dominance and violence

3.4. Conceptualization and measurement of dominance

The empirical studies reviewed led to the following hypotheses:

5.1. The International Dating Violence Study

5.3. Questionnaire administration

5.4. Measure of dominance

5.5. Measures of partner violence

5.5.1. The CTS2

5.5.2. Mutuality types

5.5.3. Physical assault

5.6. Limited Disclosure Scale

5.7. Gender symmetry

5.8. Demographic characteristics

5.8.3. Relationship length

6. Prevalence of partner violence

6.1. Assault rates

6.1.1. Overall assault rate

6.1.2. Severe assault

6.1.3. Ratio of male to female assaults

6.2. Bidirectionality and asymmetry

6.2.1. Prevalence of bidirectional and asymmetric violence

Fig. 1. Percent of violent couples in each mutuality type.

6.2.2. National setting differences

6.2.3. Age and partner violence

6.2.4. Symmetry in severity

7. Dominance by males and females

7.1. Validity of the dominance scale

8. Relation of dominance to partner violence

8.1. Dominance and overall violence

8.1.1. Dominance by men and Male-Only violence

8.1.2. Dominance by men and Female-Only violence

8.1.3. Dominance by men and bidirectional violence

8.1.4. Dominance by women and Male-Only violence

8.1.5. Dominance by women and Female-Only violence

8.1.6. Dominance by women and bidirectional violence

8.2. Predicted probability graphs

8.3. Dominance and severe violence

9.1.1. Student sample

9.1.2. Clinical versus community sample

9.1.3. Cross-sectional data

9.2. Theoretical and methodological implications

9.2.1. Theoretical implications

9.2.2. Risk factors versus one-to-one causes

9.2.3. Types of violence

9.3. Policy and practice implications

Вам также может понравиться